Fault Detection and Isolation of a Wind Turbine

Abstract

In this paper, several methods used for detection and isolation of different types of faults which can affect a wind turbine are presented. These methods can be divided into two categories: model-based methods (designing a state-space estimator is considered) and methods based on the wind turbine’s signals analysis (in both frequency and time domains). A benchmark model implemented in Matlab and Simulink of a real wind turbine (Odgaard, et al., 2009) is used to test these fault detection and isolation methods. The faults considered in this model affect some of the sensors and the actuators of the wind turbine. The developed methods are implemented as Simulink blocks and placed in the wind turbine’s benchmark model, allowing an online detection and isolation of the considered faults. The obtained results are then analysed.